Integrate the chatbot with the SAP Intelligent RPA

1. Open the project created in group Automate a Procurement Business Process Using SAP Intelligent RPA.

2. Go to the Overview tab and click Generate Package.

   

3. From the pop up, check the Name and Version Number and click Generate Package.

   

4. Navigate to the Packages tab in the Cloud Factory to view the generated package. Search for the Orders Management package.

5. Click the drop-down. From the line-item generated, click … and then click Release to release the package.

   

6. In the Release Package pop-up, click Release. The package will be in Released status.

   

This step involves creating a trigger of type API and an API Key in your environment. A trigger of type API opens a dedicated endpoint that allows an external application to start the execution of an automation or scenario in a specified deployed package via an HTTP POST call. As such, an API trigger is always attached to one automation or scenario in one deployed package. A deployed package is bound to an environment, which means that using an API trigger will guarantee job execution on one of the agents in the environment of the deployed package.

1. Click … against the released version and then click Create Trigger.

   

2. In the Create Trigger pop-up, select an environment and click Next.

   

3. In Step 2 Deploy, package will be deployed automatically and in Step 3 Set Environment Variables, the variables will be set, if any. In Step 4 Select a trigger type, select the radio button API and click Next.

   

   The Create API trigger pop-up is displayed.

4. In the Create API trigger pop-up, provide the following information:

   Field	Value
   Name	API_CAI_Call
   Description	An API triggered from CAI chatbot
   Execute	Collect Order Details
   Priority	Medium
   Job Expiration	60 Minutes

   

   Click Create. The API trigger information pop-up is generated.

5. In the API trigger information pop-up, click Copy and paste it in a notepad. Click Close. The newly generated trigger is visible in your Environment.

   

6. Click API Keys in your Environment and then click Add API Key. The Generate New API Key pop-up is displayed. In the Generate New API Key pop-up, input the Name as myAPIKey and click Next.

   

7. Activate environment_read, trigger_read and trigger_execute and click Next.

   

8. Review your API Key and click Add.

9. Click Copy and make sure to note down the key since you will not be able to view it again.

   

The API trigger and the API Key are created successfully.

This step involves creating a Service Key to access the SAP Intelligent RPA APIs. You will need the administrator rights of your global account in the SAP BTP Cockpit.

1. Log in to your SAP BTP Cockpit and navigate to your Subaccount.

2. Create a new Service Key by navigating to Services > Instances and Subscriptions > Instances > select your IRPA instance > Service Keys > Create.

   

3. Put the name as myServiceKey and click Create.

   

4. Click … and then click View to view the Service Key.

   

5. Click Form and copy the clientid, clientsecret and URL. You will need it in the next steps.

   

In this step, you will create a chatbot in the SAP Conversational AI and fork intents to use it’s expressions in your chatbot.
Further, you will configure the chatbot to integrate with the SAP Intelligent RPA.

1. Log in to your Conversational AI account.

2. Create a new chatbot by providing the following information and click Create.

   Option	Value
   What do you want your chatbot to do?	Perform Actions
   Select predefined skills for your bot	Greetings
   Bot name	cai-with-irpa
   Description	A chatbot to get the order statuses
   Topics	customer-support, Administration
   Data Policy	Non-personal
   Store conversation data	Store
   Bot visibility	Public

3. Fork the intent getorders.

   

4. The forked intent is now displayed in your list of intents. Click the newly created intent getorders to check the list of expressions.

5. Click Train and then navigate to Build tab.

6. Click Add Skill and input the following information in the Add Skill pop-up and click Add.

   Field Name	Value
   Name	api_call
   Type	Business

   

7. Go to the skill api_call and open the Triggers tab.

8. Add If statement. If @getorders and click Save.

   
   

9. Open the Actions tab and click New Action Group.

1. Click Connect External Service and then click CONSUME API SERVICE.

   

2. In the API Service Configuration, paste the URL you copied in Step 2.5.

3. Choose OAuth2 Authentication as authentication type. The Service Key created in Step 3 will be used in this step. Use the clientid, clientsecret and URL copied earlier. For Authorization URL, make sure to add the suffix: /oauth/token. For example: https://<yourtenant>.authentication.us10.hana.ondemand.com/oauth/token

   Input the following information:

   Field Name	Value
   Client ID	clientid copied in Step 3.5
   Client secret	clientsecret copied in Step 3.5
   Authorization URL	URL copied in Step 3.5 appended with suffix /oauth/token

   

4. Next, go to the Headers tab and click New Fields.

   

5. Input the following header information:

   Field Name	Value
   Key	irpa-api-key
   Value	API Key copied from Step 2.9

   

6. Now, go to the Body tab and write the following JSON body.

   JSON

   Copy

   {
       "invocationContext": {
       "conversationId": "{{conversation_id}}"
       },  
       "input": {}
   }
   

   Make a note of the conversationId variable used to pass the chatbot’s conversation_id. You will be referring to this variable later during the tutorial.

7. Click Save to save changes.

   

Now, the chatbot is ready to connect via API Service Call.

You must generate Designtime APIs token in SAP CAI to be able to authenticate the APIs successfully.

1. Within the chatbot, click Settings

   
   and then click Tokens.
   

2. Copy the Developer token under Bot token and paste it in a notepad.

   

3. Click Generate to generate a Designtime API token.

   

4. Select Client Credentials radio button and click Next.

   

5. Note down Auth URL, Client ID and Client Secret from the generated token. These credentials are valid for 12 hours only.

   

The Developer token and Designtime APIs token information will be used in the next step to create Notifiers.

In this step, you will create Notifiers.

You will create three Notifiers for notifying Start, Success and Failure status while sending message to chatbot in CAI. The Start notifier will notify that the automation started. The Success notifier will notify that the automation ended successfully and results are downloaded in Excel with Orders and their Status. In case the automation fails, then the Failure notifier will send the message to chatbot with error details.

1. Open your SAP Intelligent RPA Factory and navigate to Environments. Go to your environment.

   

2. From the trigger API_CAI_Call which you created in Step 2.5, click More options

   
   and click Add Notifier.
   

3. Select CAI radio button and click Next.

   

4. In the Create CAI Notifier pop-up, fill in the following details and click Create:

   Field Name	Value
   Notify on	Start
   Name	Starting notifier
   Description	Indicates that the automation has started
   Conversation ID	${invocation_context.conversationId}
   Token	Developer token from Step 6.2.
   Auth URL	Designtime APIs token Auth URL from Step 6.5
   Client ID	Designtime APIs token Client ID from Step 6.5
   Client Secret	Designtime APIs token Client Secret from Step 6.5
   What do you want to do?	Send a message
   Delay in second	0
   Message content	Hello from IRPA bot! Automation started.

   

   Make sure the Conversation ID variable conversationId is the same as defined in Step 5.6. In case, the chatbot’s conversation_id is passed in a different variable in Step 5.6 then, use the same variable in the Notifier. If the variables do not match, then the integration will fail because the Notifier will not be able to send the message to incorrect chatbot’s conversation ID.

5. Repeat the steps 2 to 4 to add Notifiers with the following details:

   Field Name	Value
   Notify on	Success
   Name	Success notifier
   Description	Indicates that the automation has finished successfully
   Conversation ID	${invocation_context.conversationId}
   Token	Developer token
   Auth URL	Designtime APIs token Auth URL
   Client ID	Designtime APIs token Client ID
   Client Secret	Designtime APIs token Client Secret
   What do you want to do?	Send a message
   Delay in second	0
   Message content	Automation completed! Your order details are downloaded in Excel.

   Field Name	Value
   Notify on	Failure
   Name	Failure notifier
   Description	Indicates that the automation has failed
   Conversation ID	${invocation_context.conversationId}
   Token	Developer token
   Auth URL	Designtime APIs token Auth URL
   Client ID	Designtime APIs token Client ID
   Client Secret	Designtime APIs token Client Secret
   What do you want to do?	Send a message
   Delay in second	0
   Message content	Automation failed with error: ${error.details}

   
   

You have successfully created the Notifiers.

This step involves the creation of an SAP Conversational AI Web Client in your CAI chatbot.

1. Open your SAP Conversational AI chatbot.

2. Click Connect tab.

3. Click on SAP Conversational AI Web Client.

   

4. Put CAI with IRPA in Channel Name and click Create.

   

5. Click on the Web Client standalone link to open the chatbot in browser.

6. The chatbot opens in a new window. You can now start your automation by typing Start my automation. You can choose any expression maintained in getorders Intent.

   Make sure your Desktop Agent is in Background (Unattended) mode.

   If the execution is successful, the chatbot displays the following output:

   

   Otherwise, if the execution fails, the chatbot displays the following error details:

   

7. Check the automation job status in Cloud Factory > Monitoring > Jobs. Look out for your job and click on it to view details. This step is helpful to know about the error details of a failed integration.